DEV Community


Posted on • Originally published at Medium on

Transitioning to Data Science — My journey so far

Since you’re little, you are constantly asked the question: “What do you want to be when you’re a grown up?”. I remember changing my mind through the years, listing a lot of different professions. But all of them had one thing in common: they all involved answering questions. I always loved trying to find answers to how things worked, or how two things that I saw were connected. I became interested in biology when I was 16, due to a very good teacher I had in high school, and soon after, I decided to study biological sciences to become eventually a full-time researcher.

Through my PhD, I encounter wonderful things: trying to dissect a complicated phenomenon by understanding it and converting it into simpler questions, gather information about it to establish a background, planning and performing experiments to answer those questions, and finally interpreting and communicating our results in research papers, or orally at several national and international conferences were things fulfilling the reason why I wanted to be a biologist in the first place. As well, the possibility of doing an internship in US allowed me to meet an amazing research group that help me shape my character as a researcher and also as a person. However, everything was not bright and I also encounter plenty tough times and a lot of disappointment that dampened my enthusiasm over academia.

All of the professions I could think of when I was growing up involved one thing: Answering questions about how things worked or were connected

By the time I defended my dissertation, all the certainties with which I had started were gone. I was beginning to wonder if academia was the right setting in which I wanted to develop my career in a way that there was a balance between personal life and work. I knew I was not alone. I had long talk with colleagues and friends who were going through the same issue. Nevertheless, I decided to go to on with a postdoctoral position while I figured out where in the system I fitted in. I received later an job offer from a US colleague that was moving back to Germany, which I took still exploring the future direction of my career.

As a postdoc, I really enjoyed managing projects as well as mentoring and supervising students; it’s something which I personally find rewarding for getting not only to teach what I know, but also to learn from the different experiences and people. However, the lack of collaborative spirit to carry out a mutual objective or the absence of rigurosity to work along with the publish or perish culture pushed me further away from academia.

In the search for the biggest answer I had to dealt with till then, “Do I want this for myself?”, I started reading about bioinformatics and big data. I liked very much the concept of it being an interdisciplinary field so out of curiosity, I enrolled in “Finding Hidden Messages in DNA” course in Coursera, part of a Bioinformatics Specialization offer by University of California San Diego. It’s a well organised course were you can opt whether you want to go for the coding track or stay with theoretical knowledge about the algorithms. I chose the first one. Due to my background, the biological concepts involved were familiar to me. The coding was another story! It was my first approach to Python and I was already struggling with slicing a string, let alone defining a function! Anyway, I was able to finish the course and to my surprise I discovered something: I loved coding in Python!! I enrolled in the next course Genome Sequencing. I slowly improved my skills in Python but by the end of the course, the algorithms were overwhelming and I realised that my trial-error strategy with coding will not work for long.

I decided then to take a step back and learn some basic concepts enrolling in Genomics Data Science specialization, offered by John Hopkins University and designed to take the students from introducing basic concepts to applying statistics and algorithms, while learning python, the R package Bioconductor and command line tools as well as Galaxy software. Once I was done with this, I returned to complete the Bioinformatic Specialization, which provided me with a variety of innovative concepts and gave me tools to understand how complex algorithms work and how they can be applied to biological questions.

Though I enjoyed learning how to code and I had strengthened my python skills, I was still not completely able to separate myself from academia. Maybe because it was my comfort zone; perhaps due to the fact that it is easy to apply bioinformatics to research. However, there was still something unsettling about the idea of me staying. It was not until I completely immerse myself into investigating the applications of big data and bioinformatics, that I discovered the unlimited potential of data science_._ The more I read, the more I became interested in that field. I started to read every blog I could find concerning its applications, the skills involved as well as the experiences of people working in it.

The more I read about the data science field, the more I became interested. I started to read every blog that I could find related to its applications and the experience of people working in it.

I decided to start building my future towards that direction, working in projects and doing courses in the evenings and weekends. I enrolled in Applied Data Science with Python (offered by University of Michigan) as well as in Statistics with R (offered by Duke University) specializations. The first one, it is a very useful and comprehensive five-courses specialization, where you can have an overview and a first approach to working with pandas, matplotlib, scikit-learn as well as to machine learning and text mining basic concepts. The second one covers from basic to Bayesian statistics while doing exercise and peer-review projects in R, and it is completely worth taking as the teacher is amazingly clear in explaining all the theory. I also enrolled in Andrew Ng’s machine learning course, which as almost every post or blog that I found commented, it is a must!

Something very useful for me was listening to the first KaggleConf that took place in March 2018 aimed to people searching their first job in data science. The talks and discussion gave tips about creating a compelling portfolio to show your projects, tailoring your CV, finding opportunities and preparing for a potential interview.

So, I finally made up my mind: I’m leaving academia and most importantly, I found out that data science might be the right setting to apply my skills, which I’m trying to enhance, in a way that better satisfy my vision on life and work.

The next step after taking this decision and finishing online specializations was to start my own data science projects that you can read about it here.

Top comments (0)